Two recent events appear unrelated but are connected misuses of science that can lead to disastrous policy implications. The first was a press conference reporting the results of a National Institutes of Health study showing that infants who were in day care were more likely to be aggressive and violent in kindergarten than children who were not in day care. The second event was a report in The Journal of the American Medical Association concluding that cocaine use during pregnancy does not produce deficits in children and is like cigarette smoking or other risk factors.

The problem with the day care report is that a week later, a second report from other scientists working on the same project told us that there is more to this story. It turns out that the rates of violent behavior of the children in day care were the same as the national norms and that other factors, such as family factors, were better predictors of violent behavior in kindergarten than whether or not the child was in day care.

The problem with the cocaine study is that it is a review of mostly studies of infants and young children. The long-term outcome studies are still in progress and there are many examples of conditions that effect children (e.g. school failure, attention deficit hyperactivity disorder, conduct disorders, autism) that are not detected until older ages.

What the day care and cocaine reports have in common is the premature release of scientific information. In the case of the day care report, the press conference was called before the study was published in a scientific journal, by passing the peer review process. And given that other authors of the study took exception with the press conference report, one wonders what version of the data would be published.

The cocaine report was published, but it was a review. In contrast to a meta-analysis that provides quantification and statistical analysis of studies, this was simply one group of scientists' interpretation of other scientists' interpretation of their own work. One of the problems with the cocaine literature is that there are probably as many reviews as there are actual studies and because the findings are so lean, scientists take the same body of literature and come up with opposite findings. Until there is an adequate scientific basis, there is no justification for drawing conclusions, frustrating as that may be.

How does this happen? To paraphrase a former professor of mine, "amoebas don't study amoebas, stars don't study stars but people study people. And when they do, they bring all of their foibles and prejudices to the study of human development," These reports were written by well-respected scientist who are passionate about their beliefs. Passion is a vital force that fuels our work. But we cross the line when we let that passion lead to the misuse of science to further one's own political agenda.

Especially at a time when our society is terrified about violence, it is irresponsible to scare the thousands of families that have to confront the problem of day care. These families are already torn by the emotional and financial conflict that this decision brings up. Now they are told that they are responsible for the next generation of "serial killers" based on unsupported conclusions from an unpublished single study.

In addition, there is legitimate concern about the many women and children that have been wronged by prejudice against drug-using pregnant women. The criminalization of these women and the loss of many of their children to the foster care system are only two of the wrongs that need to be righted. The recent Supreme Court ruling that it is illegal for birthing hospitals to provide drug screen results to law enforcement agencies is a step in the right direction. However, the premature conclusion that cocaine use during pregnancy is not harmful has the potential to undo more than 20 years of developing treatment programs for these mothers and their children.

And what are the policy implications of saying that cocaine and cigarettes have similar effects? Cigarette smoking has already been linked to low birth weight, there is some evidence relating it to sudden infant death syndrome and there are a few reports relating prenatal cigarette smoking to cognitive and behavioral problems. Approximately 20 percent of women smoke during pregnancy. That translates into almost 800,000 infants born in the United States per year with prenatal nicotine exposure. Are we going to take an already overburdened child welfare system and report these mothers to social services like we do with cocaine using mothers?

If we accept the day care and cocaine reports, we would conclude that day care causes violent behavior in kindergarten while cocaine does not. Therefore, day care must be worse for children than prenatal cocaine exposure.

Barry Lester, Ph.D., is professor of psychiatry and human behavior and pediatrics at Brown and director of the Infant Development Center at Women & Infants Hospital in Providence, R.I. He can be contacted at (401) 453-7640.

Poor children benefit from established public preschool and school-based intervention programs, according to a 15-year follow-up study of 1,539 children enrolled in 25 Chicago-based early childhood programs. Researcher Arthur J. Reynolds, Ph.D., of the University of Wisconsin at Madison, and colleagues conducted the nonrandomonized, matched-group cohort study to evaluate the long-term benefits of public early childhood programs, especially in the areas of educational attainment and crime reduction.

Outcomes were assessed using high school graduation and dropout rates, juvenile arrest records, grade retention and special education placement data. After adjusting for covariates, the researchers found that preschool children enrolled in the intervention for one to two years experienced higher graduation rates (p=0.01) and completed more years of education (p=0.03) than the comparison group. These children also experienced a significant decrease in arrest and dropout rates.

"Participation in an established early childhood intervention for low-income children was associated with better educational and social outcomes up to age 20 years," the authors conclude. "These findings are among the strongest evidence that established programs administered through public schools can promote children's long-term success."

Reynolds AJ, Temple JA, Robertson DL, et al.: Long-term effects of an early childhood intervention on educational achievement and juvenile arrest. The Journal of the American Medical Association 2001; 285 (18):2339-2346. For reprints, contact Arthur J. Reynolds, Ph.D., Waisman Center on Mental Retardation and Human Development, University of Wisconsin-Madison, 1500 Highland Ave., Madison, WI 53705; e-mail: ajreynol@facstaff.wisc.edu.